To investigate the use of information content (IC) as a screening tool for identifying rare disease candidates using SNOMED CT, particularly within the context of the Singapore health system's unique challenges.
Key Findings:
IC effectively distinguished RD patient profiles from the first clinical encounter, suggesting a need for early intervention.
The proposed method surfaced 71 underdiagnosed rare diseases, 57 of genetic origin, highlighting gaps in current diagnostic practices.
Achieved 20% precision starting from 3 encounters with an IC threshold of 8.17, indicating a feasible approach for clinical implementation.
Interpretation:
The study demonstrates the potential of using information-theoretic metrics in EHR for rare disease screening, highlighting a novel approach that could significantly improve identification and diagnosis in clinical settings.
Limitations:
The study is limited to the Singapore health system and may not be generalizable to other regions, which could affect the applicability of the findings.
Potential biases in the dataset due to incomplete health records or misclassification may impact the reliability of the identified rare disease cases.
Conclusion:
This is the first study to apply information-theoretic metrics to EHR for rare disease screening, indicating a promising direction for improving diagnostic processes in health systems.
Microdroplets formed during electrospray ionization may trigger chemical reactions that help explain a substantial portion of the “dark metabolome” – though some researchers question their relevance under typical metabolomics conditions
